Downhole Physical Property Logging of the Blötberget Iron Deposit, Bergslagen, Sweden Geofysisk Borrhålsloggning I Apatitjärnmalmer, Norra Bergslagen

Independent Project at the Department of Earth Sciences Självständigt arbete vid Institutionen för geovetenskaper 2017: 32

Downhole Physical Property Logging of the Blötberget Iron Deposit, Bergslagen, Sweden Geofysisk borrhålsloggning i apatitjärnmalmer, norra Bergslagen

Philip Johansson

DEPARTMENT OF

EARTH SCIENCES

INSTITUTIONEN FÖR

GEOVETENSKAPER

Independent Project at the Department of Earth Sciences Självständigt arbete vid Institutionen för geovetenskaper 2017: 32

Downhole Physical Property Logging of the Blötberget Iron Deposit, Bergslagen, Sweden Geofysisk borrhålsloggning i apatitjärnmalmer, norra Bergslagen

Philip Johansson

Downhole Physical Property Logging of the Blötberget Iron Deposit, Bergslagen, Sweden Philip Johansson

Geophysical methods are frequently applied in conjunction with exploration efforts to increase the understanding of the surveyed area. Their purpose is to determine the nature of the geophysical response of the subsurface, which can reveal the lithological and structural character. By combining geophysical measurements with the drill core data, greater clarity can be achieved about the structures and lithology of the borehole. The purpose of the project was to give the student an opportunity to discover borehole logging operations and to have a greater understanding of the local geology, in particular the iron mineralizations in the apatite iron ore intersected by the boreholes. In order to do this, the student participated in performing a geophysical borehole survey and analyzing the results. These were combined with a drill core log in order to cross plot the results and increase understanding.

Key words: Borehole logging, resistivity logging, magnetic susceptibility, magnetite, Bergslagen

Independent Project in Earth Science, 1GV029, 15 credits, 2017 Supervisor: Alireza Malehmir Department of Earth Sciences, Uppsala University, Villavägen 16, SE-752 36 Uppsala (www.geo.uu.se)

The whole document is available at www.diva-portal.org

Sammanfattning

Geofysisk borrhålsloggning i apatitjärnmalmer, norra Bergslagen Philip Johansson

Geofysiska metoder används ofta i samband med prospektering för att öka förståelsen av området. Utförda från ytan ger de en relativt god tolkning av vad som kan finnas på djupet och är även kostnadseffektiva jämfört med provborrning. Borrhålsloggning sker däremot efter att själva hålet borrats och ändamålet är ofta att utöka förståelsen om området omedelbart kring det loggade hålet. Genom att kombinera geofysisk fältdata med tolkning av borrkärnan kan man erhålla en ökad förståelse för borrhålets strukturer och litologi. Syftet med det här projektet var att utöka studentens förståelse inom borrhålsloggning, samt att avgöra hur relevant metoden är för att identifiera järnmineraliseringar i apatit- järnmalmen som kännetecknar norra Bergslagen.

Nyckelord: Borrhålsloggning, geofysik, apatitjärnmalm, Bergslagen

Självständigt arbete i geovetenskap, 1GV029, 15 hp, 2017 Handledare: Alireza Malehmir Institutionen för geovetenskaper, Uppsala universitet, Villavägen 16, 752 36 Uppsala (www.geo.uu.se)

Hela publikationen finns tillgänglig på www.diva-portal.org

Table of Contents

1. Background ...... 1 1.1 Geological setting...... 1 2. Method ...... 2 2.1 Setup ...... 3 2.2 Temperature & resistivity ...... 4 2.3 Magnetic susceptibility probe ...... 4 2.4 Sonic probe ...... 4 2.5 Geoelectric probe ...... 4 2.6 Background gamma radiation ...... 5 3. Results ...... 6 4. Discussion ...... 10 5. Conclusions ...... 12 Reference list...... 12 Appendix ...... 14

1. Background The area around Ludvika has been mined for iron since the 16th century. In the 1970’s the mine was shut down due to falling iron prices and international competition. Higher prices have, as of a few years ago, stimulated a boom in exploration companies worldwide. The one consequence being that old deposits become re-evaluated for potential mining. By going through old mining logs and performing new surveys, companies strive to more accurately determine the grade and magnitude of ore viable for extraction (the so called ore evaluation). In Ludvika, this means that the old iron mines’ archives have been evaluated and new drill cores have been retrieved from the field (Nordic Iron Ore, 2015). 1.1 Geological setting Located in southern Dalecarlia, Ludvika is at a low point in the topography, surrounded by forest dominated by evergreens. The soil is almost without exception of glacial till, owing to the last ice age. Because of the location in central Sweden, the area shows similar mineralization as the rest of Bergslagen. The most significant rock types are different metavolcanics, in which the iron mineralization is mainly found. Appearing as both hematite and as magnetite, the ore body in Blötberget contains a mix of the two as well as localized parts where they are found isolated. The inferred magnitude of extractable ore at Blötberget (see figure 1) is currently 10,2Mt, with an iron percentage of around 42,9% (Lindholm, 2011). The remaining 57,1% are non-iron bearing minerals as well as minerals containing iron but including impurities complicating practical extraction. The surveyed hole in my study went to a depth of 480 metres, and was Figure 1. Map of the general area in which the steeply oriented to the N-S direction. The survey took place, shown are Blötberget mountain as well as the town of Ludvika. Source: “GSD- bedrock in the area is mainly a mixture of Fastighetskartan”, 1:10 000 © Lantmäteriet. intrusive and extrusive meta-granites and meta-diorites, with light, felsic to intermediary minerals predominating. The formations emplaced approximately 1,9Ga in a volcanic back-arc setting (Allen, et al., 1996) and the iron mineralization is mingled with apatite and is recently claimed to be of the “Kiruna type” apatite-iron-oxide ore. The apatite frequently occurs as fluorapatite, and contains trace amounts of rare earth elements (REE) (Frietsch & Perdahl, 1995). The main mineralization

1 includes both magnetite and hematite. The ore occurs in gently dipping massive bands with a thickness ranging from a few metres to about 15m. The mineralized body is open at depth beyond 700 metres, which is where the deepest borehole ends (Nordic Iron Ore, 2015).

Figure 2. Petrological map of the area. Red and orange areas represent massive granite respective rhyolite deposits. Source: “Berggrundskarta”, 1:5000, © SGU.

2. Method As the borehole was 76mm in diameter, the process of slimhole logging was used. Various long, thin probes were employed for different measurements. The depth of the borehole was estimated to be 480m. The casing of the borehole went to a depth of 5m.The method mainly focused on measuring the various physical properties of the formations intersected by the borehole, correlating the readings with the depth and consulting Nordic Iron Ore about the cores. In order to plot the data from the loggings, MATLAB software was used in conjunction with scripts from an earlier logging of the same character. Later R was used to plot and correlate the data. The probes were from Robertson Geologging Ltd. and included instruments for measuring magnetic susceptibility, natural gamma radiation, borehole fluid temperature, acoustic properties, and electrical conductivity. These properties were measured by four probes, all of which had the ability to measure the natural gamma radiation. A petroleum generator powered the winch as well as the logging equipment. Nordic Iron Ore supplied a core log. Knowing the general stratigraphy of the borehole, the data was combined with reference values for

2 the physical properties of the rocks contained therein. These results were a reference against which to test the readings from the field survey. Measurements were further statistically cross-correlated in order to see their dependency on each other. The principal methods of which this was done was with the Spearman’s rank coefficient, which shows the tendency for monotonic correlation between two variables, and the Pearson correlation coefficient, which ranks linear relationships. The script used for this can be found in the appendix. These coefficients were used when determining which parameter pairs to be used in the cross-plots, with the required threshold being set to ±0,50. Only pairs who exceeded the threshold on both parameters were used when cross-plotting, this was done in order to limit the total amount of plots that had to be drawn. The data was rather noisy, and in order to minimize the effects this had on the graphs; it was reduced using a one dimensional median filter. The plots were then created using a density scatter function, which colours the plot area by rate of occurrence similar to a histogram.

Figure 3. Probes from left to right, temperature, magnetic susceptibility, geoelectric and triple full- waveform sonic probes respectively. Photo by Philip Johansson 2.1 Setup A trailer with the winch was backed up and aligned with the borehole, above which a tripod with a pulley was assembled. Before any real probes were tried in the hole, a dummy was winched to the bottom and up again. This was to make certain that the hole could be accessed smoothly in order for the expensive probe not to get stuck. Assembling the probes with the winching equipment required about fifteen minutes. The procedure required the probe to be lowered to the bottom of the borehole, to then be slowly raised up while taking measurements. The maximum speed that the probe could be raised while logging differed between methods, but all fell within 2 m/ min to 5 m/ min.

2.2 Temperature & resistivity Borehole fluid temperature is ideally measured first; this is to ensure that the fluid was undisturbed, and not vertically mixed. In our case, and for organizational reasons, this came as the first measurement after a night of rest. This also meant that the measurements were registered as the probe was descending, as opposed to the other probes, where the survey was done on the ascent. The probe simultaneously sampled the resistivity of the fluid, using an electric current. This showed up as conductivity in the results. (Knödel, Lange and Voigt, 2008)(Deganwy & Conwy, n.d.). 2.3 Magnetic susceptibility probe The magnetic susceptibility of a rock, is the level of magnetization it can attain when affected by an external magnetic field. The probe therefore includes an induction system containing a set of coils. Through which a current is run, creating a weak magnetic field. Materials affected by the field, such as those being para- or diamagnetic, will create a field with the same, or inverted polarity as the inducing field, which then shows up as either constructive, or destructive interference (Knödel, Lange and Voigt, 2008). The same is true for ferro- or ferrimagnetic materials, as the inherent magnetic field of those affect the induced field (Beckman, 2015).The range of detection amounts to 16 cm starting from the probes’ casing, giving a penetration depth of between 5-7 cm to 16 cm depending on how centered the probe is in the hole(Deganwy & Conwy, n.d.). The method is comparably fast with a probe speed of up to 3 m per minute. Higher speeds could be used but with the risk of impeding the measurements. 2.4 Sonic probe The full waveform sonic probe measures the waveform of a transmitted mechanical signal, which is used to estimate rock velocity from the transit time between the source and receivers (a triple sonic probe was used). To accomplish this, the probe is fitted with centralizers meant to keep it equidistant from the hole walls. During the survey an emitter on one end was set to generate waves in the KHz frequency while the probe was raised at a speed of 5 m/ minute. Three receivers in the other end measured the waveform and transit times of these, and the difference in arrival times is used to calculate the velocity of the surrounding rock (Timur & Toksöz, 1985). This meant that cavities, fissures and faults are typically characterized by an attenuation of the signal and potentially a decrease in the seismic velocity. 2.5 Geoelectric probe The geoelectric probe utilizes two types of measurements, the first is passive and measures the borehole fluid’s self-potential (SP), the other is active, and measures an induced result. The SP, is the spontaneous current that occurs between naturally occurring cathodes and anodes (Dentith & Mudge, 2014). This happens quite frequently in bodies

4 containing, among others, magnetite mineralizations, where the electric potential varies greatly between the ore and the county rock. A second measurement is the resistivity of the rocks surrounding the borehole, and does this by having a four electrode array emitting a current which passes through the borehole fluid and into the walls, returning to the probe cable where the signal is converted from to formation resistivity (Knödel, Lange and Voigt, 2008). Both of these are useful when determining borehole water quality, where the resistivity and self-potential usually increases with the amount of free ions available. These being an indication of dissolved substances. 2.6 Background gamma radiation All of the probes include a natural gamma ray measuring device for calibration purposes. The major emitters of gamma radiation found in nature are mainly isotopes of thorium, uranium and potassium, and the daughter products the decay of these produces. These exist in trace amounts in most mineralization, and the response induced in the probe gives an indication of their concentration (Timur & Toksöz, 1985). Since there was no way of knowing the source of the gamma radiation, an assumption had to be made that these elements existed in reasonable ratios, and which minerals they were likely to exist in. It was, however, useful in the sense that variations in radiation could be used to determine bed thickness, potential fracture zones, and to make estimates about borehole lithology when correlated with core logs.

3. Results The results were obtained as raw data points in .txt format. These data points were plotted using MATLAB software and the graph in Figure 4 is the result.

Figure 4. Downhole logging results color-coded based on the core logs. Borehole casing is noticeable at 0 m by the magnetic susceptibility signature. The reference values for the most commonly occurring minerals in the rocks from the drill core log are found in table 1. These values were used as an indication of the expected seismic velocities of the core rocks. Reference values for the electrical resistivity of the most commonly occurring rocks are found in table 2. Magnetic susceptibility data for general types of rocks is located in Figure 6. The correlation coefficients for the various data can be found in tables 4 and 5, and the cross-correlated graphs are found in Figures 5 and 6.

Table 1. Densities and chemical compositions for the most frequently occurring minerals in the borehole (Holmgren, 1993).

Mineral Chemical composition Density (g/cm^3)

Biotite K(Mg, Fe)3(AlSi3O10)(F, OH)2 2.80-3.30

Beryl Be3Al2(Si O3)6 2.70-2.90

Calcite CaCO3 2.7

Chlorite (Mg,Fe)3(Si,Al)4O10(OH)2·(Mg,Fe)3(OH)6 2.60-3.30 3+ Epidote {Ca2}{Al2Fe }[O|OH|SiO4|Si2O7] 3.30-3.90 2+ X3Y2(SiO4)3 X = Ca, Mn, Fe , Mg (Zn, Y, Na); Y = Al, Cr, F Garnet (Ti3+4+, V3+, Fe2+) 3.51-4.32

Hornblende Ca2(Mg, Fe, Al)5 (Al, Si)8O22(OH)2 3.0-3.4 2+ 3+ Magnetite Fe Fe 2O4 5.20

Muscovite KAl2(AlSi3O10)(F,OH)2 2.80-2.90

Olivine (Mg, Fe)2SiO4 3.3-3.4 Potassium Felds KAlSi2O6 / NaAlSiO4 2.47-2.63

Plagioclase NaAlSi3O8 – CaAl2Si2O8 2.62-2.76 Pyroxene group - 3.25-3.55

Quartz SiO4 2.65

Pyrrhotite Fe1-xS (x = 0 to 0.2) 4.58-4.65

(Ca,K,Na)(Al,Fe,Li,Mg,Mn)3(Al,Cr, Fe,V)6 Tourmaline (BO3)3(Si,Al,B)6O18(OH,F)4 2.90-3.25

Table 2. Conductivity and resistivity of the expected materials in the borehole according to the drill core, note that resistivity is the inverse of conductivity (Dentith & Mudge, 2014).

Material Conductivity (S/m) Resistivity (Ωm) Native metals 10^8.0 - 10^7.0 10^-8.0 - 10^-7.0 Steel Casing 10^7.0 10^-7.0 Sulphides 10^7.0 - 10^-2.5 10^-7.0 - 10^2.5 Oxides 10^4.8 - 10^-8.3 10^-4.8 - 10^8.3 Shale/ mudrocks 10^1.3 - 10^10^-2.4 10^-1.3 - 10^2.4 Water (Saline to fresh) 10^0.8 - 10^-2.0 10^-0.8 - 10^2.0 Silicates 10^-7.0 - 10^-14.0 10^7.0 - 10^14.0 Air 10^-14.0 10^14.0

Table 3. Magnetic susceptibility values of selected borehole materials (Dentith & Mudge, 2014).

Material - Magnetic susceptibility (SI) Magnetite 10^2.0 - 10^0.0 Steel 10^1.2 - 10^0.1 Sulphide-Oxide Mineralization 10^0.7 - 10^-4.0 Magnetite rich skarn 10^0.1 - 10^-3.1 Granite 10^-2.2 - 10^-4.0 Silicates 10^-4.8 - 10^-5.0

Table 4. Spearman ranking values for the measured parameters, coefficients higher or lower than the decided threshold (±0, 50) are highlighted in yellow; the parameters of which both the Spearman’s rank as well as the Pearson coefficient met the threshold are marked in green.

DEPTH SHN TEMP LONG SP NGAM SPR TIME MSUS COND

DEPTH 1,00 X X X X X X X X X

SHN -0,43 1,00 X X X X X X X X

TEMP 0,99 -0,42 1,00 X X X X X X X

LONG -0,41 0,88 -0,40 1,00 X X X X X X

SP 0,35 -0,37 0,37 -0,48 1,00 X X X X X

NGAM -0,22 0,39 -0,23 0,37 -0,24 1,00 X X X X

SPR -0,39 0,95 -0,38 0,89 -0,41 0,40 1,00 X X X

TIME 1,00 -0,43 0,98 -0,41 0,35 -0,22 -0,39 1,00 X X

MSUS -0,10 0,09 -0,11 0,11 -0,10 0,11 0,11 -0,10 1,00 X

COND -0,91 -0,45 0,87 -0,42 0,33 -0,15 -0,41 0,14 -0,40 1,00

Table 5. Pearson linear correlation coefficients for the measured parameters, coefficients higher or lower than the decided threshold (±0, 50) are highlighted in yellow the parameters of which both the Spearman’s rank as well as the Pearson coefficient met the threshold are marked in green.

DEPTH SHN TEMP LONG SP NGAM SPR TIME MSUS COND DEPTH 1,00 X X X X X X X X X SHN -0,50 1,00 X X X X X X X X TEMP 0,77 -0,29 1,00 X X X X X X X LONG -0,45 0,86 -0,31 1,00 X X X X X X SP 0,15 -0,14 0,19 -0,19 1,00 X X X X X NGAM -0,03 0,31 0,30 0,25 0,01 1,00 X X X X SPR -0,40 0,91 -0,24 0,89 -0,13 0,29 1,00 X X X TIME -1,00 -0,50 0,77 -0,45 0,15 -0,03 -0,40 -1,00 X X MSUS 0,40 0,15 -0,25 0,15 -0,07 0,12 0,11 -0,39 1,00 X COND 0,82 -0,45 0,81 -0,47 0,21 0,33 -0,43 0,67 -0,13 1,00

Figure 5. Correlation graphs of parameters passing the threshold set for the parameters (unfiltered).

Figure 6. Correlation graphs of parameters passing the threshold set for the parameters (filtered).

4. Discussion Starting with the values from tables 1 to 3, qualitative estimates could be made about what readings the probes would give when they passed by the different lithologies. This implied that facies containing trace amounts of iron, as well as massive magnetite mineralizations would show comparatively high seismic velocities, whereas felsic facies, such as silica-rich granites, would show lower velocities owing to a comparatively lower average density. Because of the drill core also implying the existence of fracture zones, these were assumed to be areas of lower velocity due to the fractures either being filled with water, or containing pockets of air. Zones containing mica were expected to have velocities similar to-, and somewhat higher than the granite. The relative electrical conductivity of the borehole wall was approximated using table 2. Here the magnetite mineralizations as well as the borehole steel casing extending the first three meters were expected to give the highest readings, being the most conductive materials. The facies containing granite and similar felsic mineralizations such as quartz veins would therefore display comparatively lower results owing to the higher resistivity inherent in these rocks. Fractures and micro cavities were expected to display a relatively high conductivity due to these being filled with borehole fluid. It was also theorized that the water would display a higher-than-average ion content owing to leaching from the borehole minerals, with the effect of further lowering resistivity As table 4 and 5, as well as figures 5

10 and 6 show, the conductivity displayed a strong linear as well as monotonic correlation with the depth and temperature, this supports the notion that fluid at higher temperature would have a higher conductivity due to the comparatively larger solubility coefficient. Table 3 suggests that the relative magnetic susceptibility of magnetite and magnetite bearing mineralizations would give a qualitatively higher response compared to the granites, quartz veins, and micas of the borehole. As shown in figure 4, there are a few clear correlations between the graphs and the tests shown in tables 4 and 5 show extremely low correlation. As was expected, the magnetic susceptibility has a few very distinct peaks in the zones where magnetite is prevalent, but also in zones where regular granite and greywacke constitute the majority of the core. This can be related to higher than usual levels of ferric oxides existing in these. Nevertheless, it can be assumed that the higher concentrations of magnetite begin at around 350 meters downhole, this is also confirmed by the drill core log. The presence of magnetite can also be derived from the electrical log. Since the self-potential gradually increases along the increasing thickness of magnetite layers. The fluid conductivity increases almost linearly at a very low rate until around 330 meters down where it spikes before continuing to increase at the former rate. The reasons behind this are hypothesized to be related to the increased borehole fluid temperature, as this leads to an increased solubility and therefore a higher degree of free ions according to Le Chatelier’s principle (Atkins & Jones, 2010). This is supported by the fairly strong positive linear correlation between the temperature and fluid conductivity shown in table 5, as well as Figure 5. As for the fluid temperature, it increases linearly at a rate of about 1°C/100m from 100m to 375m where this increases to 3°C/100m. This increase of rate can be explained as either fracture zones leading in warmer water from another location, radioactive decay of surrounding minerals giving off heat, the geothermal gradient or a combination of the three. Seeing as the only fracture zone is about 50m up-hole the first reason seems implausible. As there are sporadic spikes in the natural gamma downhole, the second alternative is more reasonable, even though there are no major increases as the depth progresses. Except for a few outliers, the overall gamma levels seem to decrease beyond the 350m mark. The most reasonable explanation appears to be that the geothermal gradient comes into play beyond 375m. Seeing as the global average of increased temperature with depth is about 2,5°C/100m (or 25°C/km), the increase at depths shallower than 375m seem a bit low compared to the global average, while the latter rate is a lot closer to it. Looking at the gamma log, most spikes seem to congregate around areas with increased occurrence of granites, basic meta-vulcanites, micas and pegmatites. This leads to the assumption that these types of rocks would be the primary hosts of radioactive elements (mainly isotopes of potassium). The seismic velocity of a rock typically increases with density and decreases if the rock is fractured or contains cavities (Kern, et al., 2009) (Dentith & Mudge, 2014). This leads to the expectation that zones containing iron bearing minerals or minerals of mafic character would also have higher seismic velocities. This can be seen in the localized spikes around the amphibolite mineralizations as well as the clearly increasing trend as magnetite

11 becomes prevalent. There are also a few spikes around the pegmatite at about 180m, which may be attributed to the nature of the mineralization containing fewer microfractures and therefore resulting in higher velocity.

5. Conclusions With the drill core log at hand it becomes relatively easy to analyze the results from the borehole log and as such it gives supporting evidence to the analysis already performed on the core by geologists. Using the downhole surveying methods by themselves would probably lead to some difficulty when it came to the determination of the lithology, since the results only show the physical properties in the borehole. Because these could be affected by external factors, they won’t necessarily say anything about the rocks contained therein. More detailed analysis is also inhibited by the general nature of the measurements. For example, determining the source of the gamma radiation shown in some of the rocks would probably require geochemical methods or some form of spectroscopy. The problems faced with an isolated geophysical survey can be described as “trying to fit a model to reality”. All in all, the results certainly complement the existing core log and as such the method is relevant for the purpose of enhancing the understanding of boreholes made for exploration purposes, though the sample size would need to be increased if the method were to be suitable for extrapolating to the area between boreholes.

Reference list Allen, R. L., Lundström, I., Ripa, M. & Christofferson, H., (1996). ”Facies analysis of a 1.9 Ga, continental margin, back-arc, felsic caldera province with diverse Zn-Pb-Ag-(Cu-Au) sulfide and Fe oxide deposits, Bergslagen region, Sweden”. Economic Geology,vol. 91, no. 6 pp. 979- 1008. Deganwy, Conwy, (n.d.) Robertson Geologging 2000m Winch Operating Manual, Robertson Geologging Ltd. Dentith, M. & Mudge, S. T., (2014). Geophysics For The Mineral Exploration Geoscientist. Cambridge: Cambridge University Press. Frietsch, R. & Perdahl, J., (1995). Rare earth elements in apatite and magnetite in Kiruna-type iron ores .and some other iron ore types. Ore Geology Reviews, vol. 9, no. 6, pp. 489-510, Kern, H., Mengel, K., Strauss, K.W.,Ivankina, T.I., Nikitin, A.N., Kukkonen, I.T., ., (2009). Elastic wave velocities, chemistry and modal mineralogy of crustal rocks sampled by the Outokumpu scientific drill hole: Evidence from lab.measurements and modeling. Physics of the Earth and Planetary Interiors, vol. 175,no. 3-4, p. 151–166.

Knödel, K., Lange, G. and Voigt, H. (2008). Environmental Geology. Berlin, Heidelberg: Springer-Verlag Berlin Heidelberg, pp.431-474. Lindholm, T., (2011), Ludvika Mines Blötberget and Håksberg Technical assessment A Qualified Person's independent Report, Luleå: GeoVista. Timur, A. & Toksöz, M. N., (1985).Downhole Geophysical Logging. Annual Review of Earth and Planetary Sciences,vol. 13, no. 1 pp. 315-344. Internet sources

Beckman, O., (2015). Nationalencyklopedin. http://www.ne.se/uppslagsverk/encyklopedi/lång/magnetism [Accessed on May 12th 2015] Nordic Iron Ore, (2015). Projekt. http://www.nordicironore.se/sv/sidor/projekt/ [Accessed on May 15th 2015].

Appendix

Figure 7. Graph of the results

Script used in R to create the correlation tables and cross-plots.

#Import all data tables;

TC <- read.table("BB14005_GaTSf_150508_test.txt", dec=".", header = TRUE);

MAG <- read.table("BB12004_MagGa_150505_test.txt", dec=".", header = TRUE);

EIT <- read.table("BB14005_ElTGa_150507_test.txt", dec=".", header = TRUE);

#Calculate monotonic and linear relationships;

CondSP <- round((cor(TC, use = "everything", method = "spearman")), 2);

CondPe<- round((cor(TC, use = "everything", method = "pearson")), 2); write.table(CondSP, file = "Spearman_Cond-Temp.txt", col.names = TRUE, row.names = TRUE, sep = "\t"); write.table(CondPe, file = "Pearson_Cond-Temp.txt", col.names = TRUE, row.names = TRUE, sep = "\t");

MagSP <- round((cor(MAG, use = "everything", method = "spearman")), 2);

MagPe <- round((cor(MAG, use = "everything", method = "pearson")), 2); print(MagSP); print(MagPe); print(cor(EIT, use = "everything", method = "spearman")); print(cor(EIT, use = "everything", method = "pearson"));

TCMAGSP <- round((cor(TC, MAG, use = "everything", method = "spearman")), 2);

TCMAGPE <- round((cor(TC, MAG, use = "everything", method = "pearson")), 2); write.table(TCMAGSP, file = "TC-Mag-Spearman.txt", col.names = TRUE, row.names = TRUE, sep = "\t") write.table(TCMAGPE, file = "TC-Mag-Pearson.txt", col.names = TRUE, row.names = TRUE, sep = "\t")

TCEITSP <- round((cor(TC, EIT, use = "everything", method = "spearman")), 2);

TCEITPE <- round((cor(TC, EIT, use = "everything", method = "pearson")), 2); write.table(TCEITSP, file = "TC-EIT-Spearman.txt", col.names = TRUE, row.names = TRUE, sep = "\t") write.table(TCEITPE, file = "TC-EIT-Pearson.txt", col.names = TRUE, row.names = TRUE, sep = "\t")

MAGEITSP <- round((cor(MAG, EIT, use = "everything", method = "spearman")), 2);

MAGEITPE <- round((cor(MAG, EIT, use = "everything", method = "pearson")), 2);